基于属性分类加权的数据隐私保护算法  

Data Privacy Protection Algorithm Based on Attribute Classification

作　　者：欧亚湘[1] 王稳波[1] Ou Yaxiang;Wang Wenbo(Jiangxi Vocational and Techenical College of Information Application,Nanchang 0043523,China)

机构地区：[1]江西信息应用职业技术学院,南昌0043523

出　　处：《科技通报》2017年第5期180-183,191,共5页Bulletin of Science and Technology

基　　金：江西省教育厅(课题编号:JXJG-09-41-6);江西省教育厅(课题编号:JXJG-09-41-3)

摘　　要：针对目前社会随着数据共享技术的不断发展所导致的个人隐私数据不断被泄露的问题,本文提出一种新的数据隐私保护算法,该算法具备自身独特的优势,在弱泛化较为重要的属性的操作过程中,特别引入了信息增益、信息增益比率两个参数。此外,还引入了属性权重,通过这个参数可以确定在分类时数据集内对应属性的重要性,并且在匿名化操作中可以对元组数据间距离进行合理的调整。仿真实验结果表明,本文提出的数据隐私保护算法相较以往经典的数据隐私保护算法在处理数据精度以及运行效率方面更加具有优势。In view of the present society with the development of data sharing technology,as a result ofpersonal privacy data leaked problem,this paper proposes a new data privacy protection algorithm,thisalgorithm has its own unique advantages,mainly embodied in the original data by processing of privacyprotection algorithm is proposed in this paper the anonymous data in classification mining method hasbetter applicability.In the operation of weakly generalized properties,the information gain andinformation gain ratio are introduced in particular.In addition,also introduced the attribute weights,andthrough this parameter can be determined in the classification of the importance of the correspondingattributes in a data set,and can be in anonymous operation of distance between the data reasonableadjustment.The simulation results show that the proposed data privacy protection algorithm comparedwith the previous classic data privacy protection algorithm in dealing with the data accuracy and runningefficiency is more advantage.

关 键 词：共享信息 数据隐私保护 加权 泛化过程 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]